FUMBBL :: Online Blood Bowl League

Joined: Jul 01, 2004

Posted: Feb 02, 2022 - 06:39

So I have a question, and some confusion. There was a day that I really cared about CR, and while it still matters to me, it really matters very little. However I noticed this in the game I recently played. I received 105k in inducements, tied the game against a higher rated player (overall) and coach for that specific team (Ogre for me, Norse for him). While I agree I was lucky to tie, why did my rating go down and his rating go up? It seems rather counter intuitive.

https://fumbbl.com/FUMBBL.php?page=match&id=4362115

bigpoppaben Star CR 160.54 (+0.06)
Norse
Star
CR
155.42

CR 152.69 (-0.06) Emerging Star spinball
Ogre
Emerging Star
CR
146.73

Maybe I am missing something, however I believe it might have something to do with how the Ogres are rated in competitive.

wintergreen13

Joined: Apr 10, 2011

Posted: Feb 02, 2022 - 06:57

Dunno if you can make sense of this, but the formula has a few areas that might be responsible. Also, I'm not sure if this is the same formula being used now, since the division specific weight links to a page that includes Academy, which is long defunct. Ultimately, 0.06 is a tiny amount of CR, though it would be fun to know why. My guess is there are some modifiers present for matches between Norse and Ogres, or the formula does not take into account Low Cost Linemen, or there is something about [C] matches specifically that influences it.

ArrestedDevelopment

Joined: Sep 14, 2015

Posted: Feb 02, 2022 - 07:10

Looks like this is a case where the "petty cash is counted in CR calc" change of 2021 is having an effect in conjunction with BB2020 fundamentally changing the idea of "petty cash" from previous editions.

Basically - you got 105k petty cash for being the underdog in-match, due to low-cost linemen - but the CR calc sees your TV as higher because it takes the starting TV of 1660 and the spend of 100k.

This fix was put in to stop people effectively having a loss-proof method of CR gain by entering very low TV snotlings in tournaments (RRR/brawl) in early 2021, where there would be a 600tv+ gap and the CR calc would simply be so far to "expected loss" that any game lost would result in 0 CR loss, but any win would be 3.5+ CR gain.

_________________

The_Murker

Joined: Jan 30, 2011

Posted: Feb 02, 2022 - 08:13

AD.. good explanation. This makes much more sense now. (Even if unconfirmed.. it adds a possibility of sense and intention)

_________________

Join the wait-list. Watch the action. Leave the Empire. Come to Bretonnia!

spinball

Joined: Jul 01, 2004

Posted: Feb 02, 2022 - 08:33

ArrestedDevelopment wrote:

Looks like this is a case where the "petty cash is counted in CR calc" change of 2021 is having an effect in conjunction with BB2020 fundamentally changing the idea of "petty cash" from previous editions.

Basically - you got 105k petty cash for being the underdog in-match, due to low-cost linemen - but the CR calc sees your TV as higher because it takes the starting TV of 1660 and the spend of 100k.

This fix was put in to stop people effectively having a loss-proof method of CR gain by entering very low TV snotlings in tournaments (RRR/brawl) in early 2021, where there would be a 600tv+ gap and the CR calc would simply be so far to "expected loss" that any game lost would result in 0 CR loss, but any win would be 3.5+ CR gain.

I thought both teams would need to play 4 games to have it count for ranking purposes? Is this no longer a thing?

ArrestedDevelopment

Joined: Sep 14, 2015

Posted: Feb 02, 2022 - 09:15

That has not been a thing for longer than I've been registered on the site.

_________________

spinball

Joined: Jul 01, 2004

Posted: Feb 04, 2022 - 02:24

ArrestedDevelopment wrote:

That has not been a thing for longer than I've been registered on the site.

So to avoid what happened with the RRR, do you think it should be?

ArrestedDevelopment

Joined: Sep 14, 2015

Posted: Feb 04, 2022 - 07:39

Not my call to make.

And not really anyway. The coach who's teams highlighted that particular issue initially wasn't even doing it for the CR himself. Most of the people who enter tournaments don't care about it - it was just very, very apparent in the RRRs and once looked into, was having an effect site-wide (and not just on game 1-5 teams at all, it affected a very wide spectrum of games, especially since with no redraft FUMBBL had a relatively large% of quite big gap games involving teams of all races).

I don't recall the data figures exactly, but excluding games 1-5 from CR gain/loss would effectively stop a very large% of many coaches games doing anything to their rankings. The game/site/userbase has progressed long past the point when that wasn't too deleterious an approach.

_________________